China may move toward U.S. path on AI as firms poach employees

In a significant shift within the global tech landscape, a former researcher from OpenAI has taken the helm as chief AI scientist at Tencent in China, signaling a potential realignment toward the ambitious goals of artificial general intelligence (AGI). This development hints at a changing dynamic in the ongoing U.S.-China tech rivalry, particularly as Chinese firms intensify their efforts to develop advanced AI technologies. AGI, which aims to replicate or exceed human cognitive abilities, has been a long-standing objective for American companies like OpenAI, Anthropic, and Alphabet, the parent company of DeepMind. Meanwhile, Chinese enterprises have been striving to catch up, focusing on practical AI applications across various sectors, from manufacturing to consumer electronics, especially under the constraints imposed by U.S. chip export controls. Tencent's Chief AI Scientist, Yao Shunyu, who joined the company after leaving OpenAI, emphasized the need for a dedicated AGI organization in China during a recent event co-hosted with local officials. "My personal goal is that in China we should establish a long-term AGI organization," he remarked, discussing the future of AI development. Yao outlined that achieving AGI requires a solid foundation of knowledge, innovative products, and exploration of frontiers, indicating a vision that could potentially reshape the tech landscape. Highlighting his aspirations, Yao stated, "I don't think ChatGPT or Claude will be the only super-app," pointing to vast untapped potential worth trillions of dollars. He argued that the effectiveness of AI tools should take precedence over costs, advocating for smaller AI models that provide consistent performance in fundamental tasks. In contrast, the mood in the U.S. regarding AI development is becoming more cautious. Recently, Anthropic raised alarms about the risks associated with advanced models that may soon be capable of self-improvement without human intervention, urging a slowdown in new model development to mitigate societal risks. The company has consistently promoted AI safety but has faced criticism from rivals who view these warnings as a strategy to limit competition. Additionally, uncertainties surrounding U.S. immigration policies have prompted many Chinese nationals to seek job opportunities back home, even if the remuneration is lower. To attract top talent, China is increasing its investments in scientific research and aiming for breakthroughs over the next five years. Companies are fiercely competing for skilled professionals, regardless of their geographical location. Among the notable recent moves, Alibaba has reportedly recruited Hao Zhou, a researcher from Google DeepMind, to bolster its Qwen AI development. Similarly, Wu Yonghui, vice president of research at Google DeepMind, transitioned to lead research at ByteDance Seed earlier this year. Furthermore, startup Moonshot, which is behind the Kimi AI model, was founded by Yang Zhilin, a former employee of Meta AI and Google Brain, further exemplifying the cross-pollination of talent in the AI field.